Automatic extraction of contextual valence shifters

Boubel, Noémi [UCL] François, Thomas [UCL] Naets, Hubert [UCL]

In opinion mining, many linguistic structures, called contextual valence shifters, may modify the prior polarity of items. Some systems of sentiment analysis have tried to take these shifters into account, but few studies have focused on the identification of all these structures and their impact on polarized words. In this paper, we describe a method that automatically identifies contextual valence shifters. It relies on a chi-square test applied to the contingency table representing the distribution of a candidate shifter in a corpus of reviews of diverse opinions. The approach depends on two resources - a corpus of reviews and a lexicon of valence terms - to build a list of contextual valence shifters. We also introduce a set of rules used to classify the extracted contextual valence shifters according to their impact on polarized words. They make use of the Pearson residuals in contingency tables to filter candidate shifters and classify them. We show that the technique reaches an F-measure of either 0.56 or 0.66, depending on how the categories of shifters are defined.